Your First Conversation

+15 Mana ✨
OpenClaw Diagnostics Reference(DOCUMENTATION)OpenClaw Troubleshooting Guide(DOCUMENTATION)

Introduction

With OpenClaw installed, onboarded, and the dashboard running, it is time to have your first AI conversation. This lesson guides you through sending your first message via the WebChat interface and verifying that everything is working correctly using diagnostic commands. By the end, you will have confirmed that your entire setup is operational.

Key Concepts

  • WebChat is the browser-based chat interface accessible from the Control UI dashboard
  • openclaw doctor runs a comprehensive health check on your entire OpenClaw installation
  • openclaw status provides a quick snapshot of daemon health and connectivity
  • openclaw gateway status reports the connection state of each configured AI provider gateway
  • Diagnostic commands are your first line of defense when something goes wrong

Real World Context

Sending a test message after setup is similar to running a "Hello World" program after installing a new programming language. It confirms that every layer of the system works: the daemon is running, the gateway is connected, the authentication token is valid, and the AI provider is responding. Teams often include this verification step in their onboarding checklists for new developers.

Deep Dive

Sending Your First Message

Open the Control UI dashboard if it is not already running:

bash
openclaw dashboard

This starts the local web server and opens your browser to http://127.0.0.1:18789/. Navigate to the WebChat panel in the dashboard.

In the WebChat input field, type a simple message and press Enter. For example, you might type:

Hello! Can you confirm this connection is working?

The message is sent from your browser to the local daemon, which routes it through the configured gateway to the AI provider. The response streams back through the same path and appears in the WebChat panel. If you see a response, congratulations: your entire pipeline is working.

Running Diagnostic Commands

After your first conversation, it is good practice to run the diagnostic commands to get a baseline understanding of your system state.

The doctor command performs a comprehensive health check:

bash
openclaw doctor

This checks multiple aspects of your installation in sequence: Node.js version compatibility, daemon status, authentication validity, gateway connectivity, configuration file integrity, and available disk space. It reports each check as either passing or failing, similar to how a CI pipeline reports test results. Any failures include a suggested fix.

The status command gives a quick overview:

bash
openclaw status

This is a lighter check than doctor. It reports whether the daemon is running, which user is authenticated, how many channels are active, and a summary of recent activity. Think of it as a quick pulse check versus the full physical exam that doctor provides.

The gateway status command focuses specifically on provider connections:

bash
openclaw gateway status

This queries each configured gateway and reports its connection state, latency, and the models available through it. It is the most targeted diagnostic when you suspect a connectivity issue with a specific AI provider.

Interpreting Results

A healthy system looks like this when you run openclaw doctor:

bash
# Example healthy output:
# [PASS] Node.js version: v22.5.0
# [PASS] Daemon running (PID: 42351)
# [PASS] Authenticated as user@example.com
# [PASS] Gateway 'default' connected (latency: 45ms)
# [PASS] Configuration valid
# All checks passed!

Each line beginning with [PASS] indicates a successful check. If any line shows [FAIL], the doctor command includes remediation steps. For example, a failed gateway check might suggest running openclaw onboard --install-daemon to reconfigure the connection.

Troubleshooting Common Issues

If your first message does not get a response, work through the diagnostics in order:

bash
openclaw status        # Is the daemon running?
openclaw gateway status  # Are gateways connected?
openclaw doctor        # Full health check

This sequence goes from quickest to most thorough. Start with status to confirm the daemon is alive, then check gateway connectivity, and finally run the full doctor check if the issue is not obvious.

Common Pitfalls

  • Skipping verification after setup: Many users jump straight into complex tasks without confirming the basics work. A quick WebChat message and openclaw doctor run takes 30 seconds and saves hours of debugging later.
  • Ignoring [FAIL] results from openclaw doctor: Each failure includes a fix suggestion. Ignoring them often leads to intermittent issues that are harder to diagnose later.
  • Assuming a silent daemon means it is working: The daemon can be running but disconnected from gateways. Always verify with openclaw gateway status after startup.

Best Practices

  • Run openclaw doctor after every upgrade to catch configuration changes or breaking updates early.
  • Save the output of openclaw doctor when filing bug reports, as it contains all the information maintainers need to diagnose issues.
  • Use openclaw status as a quick pre-flight check before starting work each day to confirm your environment is healthy.

Summary

  • Send your first message via the WebChat panel in the Control UI dashboard at http://127.0.0.1:18789/
  • Use openclaw doctor for a comprehensive health check covering Node.js, daemon, auth, gateways, and config
  • Use openclaw status for a quick daemon and connectivity snapshot
  • Use openclaw gateway status to check individual AI provider connections
  • Always verify your installation works before diving into complex tasks
✓ Completed